Technical Field
The present application relates to the field of information processing technologies, and in particular, to a server, a client, and a video processing method.
Related Art
Identifying, in a video, an object in which people are interested can be widely used in applications such as advertisement, video surveillance. However, with the increasing video resolution and the massively growing video content, it is time-consuming and inefficient to identify full content of the video.
To improve the identification efficiency, it is feasible to process a key frame (I frame) only so as to reduce the amount of information to be processed. However, this manner does not solve the key problem for improving the efficiency of video mode identification.
In addition, the mode identification of the video is generally performed at the server. This manner is advantageous in that once the identification is completed, the identified object can be widely applied; however, it requires long time and high cost for the server to process massive video content generated over the Internet.